Research On USRP RIO Anti-jamming Frequency Hopping Communication System Based On Machine Learning

The rapid development of wireless communication technology has not only changed people’s way of life,but also changed the mode of modern and future warfare.In order to ensure the reliability of information transmission,the communication system must have the ability of anti-jamming.As the most widely used communication anti-jamming technology,frequency hopping technology has been paid more and more attention by many researchers.This paper designs the framework of antijamming adaptive frequency hopping communication system based on USRP RIO platform,and further studies and implements the anti-jamming communication system based on reinforcement learning decision and the anti-jamming communication system based on deep learning prediction.The intelligent anti-jamming technology is realized through the machine learning algorithm,which effectively enhances the anti-jamming ability of frequency hopping communication system.The specific work is as follows:(1)In order to avoid or suppress malicious interference to achieve reliable communication,this paper designs an anti-jamming adaptive frequency hopping communication system framework based on software radio platform USRP RIO.According to the basic functional requirements of the antijamming adaptive frequency hopping communication system,four main modules of user,jammer,sensing node and information processing center are designed,and the functions and physical realization of each module are described.According to the framework of adaptive frequency hopping communication system,the realization flow of system function is designed.(2)A frequency hopping spectrum state perception based on Q-learning algorithm and its update scheme are proposed to optimize the frequency band selection,and the model construction and system flow design of anti-jamming communication system based on reinforcement learning decisionmaking are realized.First of all,the double threshold energy detection is used to realize the spectrum sensing and obtain the spectrum state information.Secondly,take the spectrum status as the input,update the reward table of Q-learning,get the updated Q table through iterative training of Q-learning algorithm,and continue to monitor the spectrum status information.When the spectrum state changes,continue to update the Q table according to the above steps,otherwise keep the Q table unchanged.The system selects the optimized frequency band for frequency hopping communication according to the Q table,so as to avoid interference.Finally,through the analysis of the experimental results,it is verified that the adaptive frequency hopping system based on the proposed scheme can optimize the frequency hopping band selection and reduce the probability of interference.(3)A slot-by-slot spectrum state prediction scheme based on long short-term memory network is proposed for idle spectrum prediction,and the model construction and system flow design of antijamming communication system based on deep learning prediction are realized.First of all,the spectrum information obtained by spectrum sensing is compared with the set threshold,and the spectrum state information is represented as two states of occupied and idle.Secondly,taking the spectrum state as the input,learning and training is carried out through the long short-term memory network model,and the model parameters are used to build the long short-term memory network prediction model to predict the spectrum state slot by slot,and optimally select the idle frequency band to realize frequency hopping communication to avoid interference.Finally,through the analysis of the experimental results,it is verified that the adaptive frequency hopping system based on the proposed scheme can effectively predict the spectrum state for the selection of frequency hopping bands,help users avoid interference in advance and ensure the communication quality of users.